Not all agricultural subsidies are equal: What Slovenian
livestock farms reveal about the future of the CAP
by Imre Fertő
The EU’s Common Agricultural Policy is often judged by how much money it distributes. But new evidence from Slovenian livestock farms suggests the more important question is what kinds of support farms receive — and whether these instruments help them escape persistent inefficiency or simply make it easier to live with it.
Agricultural subsidies are rarely absent from debates about the future of European farming. For some, they are essential to maintaining rural livelihoods, food security and environmental stewardship. For others, they risk dulling incentives to innovate, restructure and become more productive. Both views can be true. The problem is that policy debates too often treat “subsidies” as a single category, when in practice the Common Agricultural Policy (CAP) is made up of very different instruments with very different effects.
Our study of Slovenian livestock farms shows why this distinction matters. Slovenia is not just a national case study; it is a useful lens through which to view a wider European problem. Many farms in Central and Eastern Europe are small, fragmented, reliant on family labour and heavily dependent on CAP support. These are precisely the conditions under which income stabilisation may be necessary, but also insufficient. When inefficiency is rooted in structural constraints, annual payments can keep farms afloat without necessarily helping them become more efficient.
The study uses micro-level Farm Accountancy Data Network data for 230 conventional Slovenian animal farms between 2014 and 2021, covering 1,807 farm-year observations. The farms include specialist dairy, specialist cattle, sheep and goat, pig and poultry, mixed livestock, and mixed crop-livestock farms. The average farm in the sample cultivates around 13 hectares, keeps about 14 livestock units and generates around €22,500 in annual output.
This structure matters. A dairy farm, a cattle farm and a sheep or goat farm do not simply represent different versions of the same production process. They use land, labour, animals, capital and intermediate inputs in different ways. Ignoring this heterogeneity risks drawing misleading conclusions from an “average farm” that does not really exist. The paper therefore applies a Bayesian dynamic stochastic frontier model that accounts both for technological heterogeneity across farm types and for the persistence of technical efficiency over time.
The central finding is straightforward: the impact of CAP support depends on the type of subsidy. Investment subsidies and other rural development-type payments are associated with higher technical efficiency in both the short and the long run. By contrast, decoupled payments and agri-environmental subsidies are associated with lower technical efficiency at both horizons. Payments for less-favoured areas show a consistently negative relationship with efficiency.
The short-run estimates make the point clearly. A one-unit increase in decoupled subsidies is associated, on average, with a 0.50% reduction in technical efficiency, while agri-environmental subsidies are associated with a 0.28% reduction. Investment subsidies and other subsidies have the opposite association, increasing technical efficiency by 0.43% and 0.39%, respectively. In economic terms, this suggests that more targeted support — especially where it helps farms modernise — may be better aligned with productivity gains than automatic income support.
The long-run results are even more important. The average long-run technical efficiency score is about 0.56, well below the short-run average of nearly 0.71. This suggests that although farms may adjust in the short term, their ability to sustain efficiency gains over time is weaker. In the long run, decoupled and agri-environmental subsidies reduce efficiency by 1.52% and 0.85% on average, while investment and other subsidies raise it by 1.31% and 1.17%.
This does not mean environmental payments are “bad”, nor that income support is unnecessary. It means they should not be expected to do all policy work at once. Agri-environmental payments may be justified by environmental benefits that are not captured by technical efficiency alone. Decoupled payments may stabilise income in a volatile sector. But if the policy objective is also to improve long-term productivity, these instruments need to be combined with measures that address the structural reasons why farms remain inefficient.
This distinction is especially important for less-favoured areas. Around 93% of farms in the sample are located in such areas, where natural or geographical constraints limit production possibilities. The study finds that an additional €100 per livestock unit in LFA support is associated with a 0.62% reduction in short-run efficiency and an 8.18% reduction in long-run efficiency. The policy interpretation should be cautious: these payments are partly compensatory, and the farms receiving them face real constraints. But the finding also points to a risk. Support that protects income without enabling adaptation may lock farms into low-efficiency trajectories.
The broader lesson for CAP reform is that persistence cuts both ways. High persistence is good for already efficient farms because it helps them sustain strong performance. But it is harmful for inefficient farms if it means that low performance is carried from one year to the next. The same subsidy can therefore have different implications depending on the farm’s starting point. This is why the paper argues that the effect of support depends not only on the instrument itself, but also on the level of technical efficiency and on whether the support reinforces or weakens persistence.
For policymakers, this means moving beyond a simple opposition between “more support” and “less support”. The more useful question is whether support is helping farms change. Area payments and eco-schemes may remain important, but they need to be paired with modernisation support, advisory services and incentives for structural adjustment. Without this integration, the CAP may stabilise farm incomes while leaving the underlying productivity problem untouched.
There are limits to what the study can claim. Like much of the stochastic frontier literature, it treats inputs as exogenous. This is reasonable for quasi-fixed inputs such as land, livestock, capital and family labour, but more problematic for flexible material inputs. The authors also note that voluntary schemes, such as investment and agri-environmental subsidies, may involve self-selection: farms that apply for and receive support may differ systematically from those that do not.
The conclusion is therefore not that one subsidy instrument should mechanically replace another. Rather, the study shows that subsidy design matters. A CAP that only compensates farms for difficult conditions may preserve rural livelihoods but fail to generate lasting productivity improvements. A CAP that combines income support with investment, advice and structural adaptation has a better chance of helping farms become more resilient.
The Slovenian case offers a wider warning. In structurally constrained farming systems, public money can either soften the consequences of inefficiency or help overcome them. The future of the CAP depends on making that distinction explicit.
Pisulewski, A., Marzec, J., Bojnec, Š. Fertő, I.
CAP subsidies, technical efficiency, and its persistence: evidence from Slovenian animal farms. Journal of Productivity Analysis 65, 27 (2026). https://doi.org/10.1007/s11123-026-00815-4
